Exploring ways to enhance K12 learning using a purpose built search engine

After a few years of helping out my son with his studies, I’ve noticed that resources provided to students often tend to be sparse. On most occasions students no longer work with complete and cohesive academic material such as textbooks, but rather snippets of digital content from various sources. I believe this tends to create a disorganized learning environment which in turn can lead to increased difficulty for students to effectively learn and succeed. To overcome this challenge, sample empirical evidence suggests that students naturally tend to dismiss partially or even entirely the provided sparse resources and instead turn to applications such as Google search to find answers. However, this in itself also introduces a new set of challenges and issues, as I’ve discussed previously in this article. These study habits suddenly expose students to potentially inaccurate and unreliable resources, counter productive user experiences, and to some degree, an unsafe learning environment. 

In hopes of eventually helping students improve their learning experience and increase their chances of success, I’ve decided to try and solve the above stated problems by developing a purpose built education search engine that uses natural language processing and Machine Learning at its core. 

The first prototype I’ve built, now live, is composed of a few key components, including an ML powered indexing system, a K12-centric data structure, a scoring algorithm, a basic ranking algorithm as well as a simple and uncluttered web interface. From a technology standpoint, one of the key differentiators lies in the indexing and scoring system, which powers the application’s semantic based search capabilities, instead of relying on simple key word aggregation. 

To tackle the problems of resource reliability and safe learning environments, the application provides a search interface to a curated list of indexed K12 educational resources. Unlike general purpose search engines, this service can only serve age and level appropriate educational content, without fail. This ensures the resources that student access and refer to are both reliable and relevant to their school year. From a user experience perspective, the indexed content also includes information about the academic subject, topic and school level, further simplifying the discovery process and ensuring students are exposed to material that will enhance their understanding of a subject. In addition, the search experience doesn’t include ads or irrelevant content, reducing the distractions and the time spent trying to find the right information. Although in a very early stage, as you can see from the recording of the live prototype below, the system is able to quickly find highly accurate content from simple natural language interactions, all presented in a short list of most relevant resources. 

This is a simple first step in my quest to leverage software applications and AI for helping students improve their educational outcomes and increase their opportunities to succeed. If you’re interested in having a conversation about this project and technology, I invite you to contact me on LinkedIn.